BGA Package Recognition via Seed Growing Pattern Analysis
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current chip mounters lack an efficient method for automatically recognizing and storing information about Ball Grid Array (BGA) packages with multiple solder ball patterns, leading to manual input errors and prolonged processing times.
Innovation Solution
A BGA package recognition apparatus and method that includes an image acquisition unit, a pattern recognition unit, and a storage unit, which analyzes image information to select seeds and perform a seed growing operation to group solder balls by pattern, size, and distance, automatically recognizing and storing information about the BGA package.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If manual input method is used to store BGA package information, then information can be stored, but it takes a long time and many errors occur
Solution Approach 1:
The patent replaces the manual mechanical input system with an automated image processing system. The pattern recognition unit automatically captures images of BGA packages and processes them to extract solder ball pattern information, eliminating the need for manual data entry and significantly reducing both time and errors.
Solution Approach 2:
The system enables self-service by allowing the BGA package information to be automatically captured and processed without human intervention. The pattern recognition unit autonomously identifies solder ball patterns and stores the information, making the system self-sufficient in acquiring package data.
2Reliability
If manual input method is used to store BGA package information, then information can be stored, but it takes a long time
Solution Approach 1:
The patent replaces the manual mechanical input system with an automated image processing system. The pattern recognition unit automatically captures images of BGA packages and processes them to extract solder ball pattern information, eliminating the need for manual data entry and significantly reducing both time and errors.
Solution Approach 2:
The system maintains continuous useful action by automatically and continuously processing BGA package images as they are supplied. The pattern recognition unit operates without interruption, continuously capturing, analyzing, and storing package information, thereby maximizing productivity while ensuring reliability.
3Reliability
If manual input method is used, then BGA package information can be stored, but many errors occur during defect inspection
Solution Approach 1:
The patent replaces the manual mechanical input system with an automated image processing system. The pattern recognition unit automatically captures images of BGA packages and processes them to extract solder ball pattern information, eliminating the need for manual data entry and significantly reducing both time and errors.
Solution Approach 2:
The system creates an accurate digital copy of the BGA package solder ball patterns through image capture and processing. This digital replica serves as a reference for defect inspection, ensuring high accuracy without requiring complex manual configuration.
4Measurement precision
If automatic recognition is implemented for BGA packages with multiple solder ball patterns, then recognition accuracy improves, but system complexity increases
Solution Approach 1:
The patent applies segmentation by dividing the complex task of recognizing multiple solder ball patterns into distinct processing stages: image capture, pattern identification, seed selection, and seed growing operation. This modular approach improves recognition accuracy while managing system complexity through structured processing.
Solution Approach 2:
The system performs partial action by focusing on identifying representative seed solder balls rather than analyzing every single solder ball in the package. The seed growing operation then extends from these seeds to classify all solder balls, reducing computational complexity while maintaining high recognition accuracy.
Data Source
AI summary
Provided is a chip mounter for recognizing a Ball Grid Array (BGA) package through a chip mounter. The chip mounter includes a BGA package recognition apparatus which includes an image acquisition unit which acquires image information of a BGA package which includes a plurality of solder balls having n patterns (n≧1), a pattern recognition unit which analyzes the image information and outputs information about the BGA package, and a storage unit which stores the information about the BGA package, wherein the pattern recognition unit recognizes the n patterns, selects n seeds respectively corresponding to the n patterns and performs a seed growing operation which groups solder balls which are continuously located adjacent to the seed and have the same pattern as the seed into the same group as the seed, with respect to each of the n seeds.


